Earlier this year, SEMrush launched the Traffic Analytics dashboard in beta. The dashboard offers a glimpse into competitors’ site performance, providing metrics not otherwise available without analytics access, like Visits, Average Visit Duration, and Bounce Rate. This is important because as you know many site performance elements are important to SEO & Content Marketing.
We wanted to test the accuracy of this new tool by using Google Analytics(GA) data for a few of our clients. Our results represent the disparity between GA data and the SEMrush data.
What is the SEMrush Traffic Analytics Dashboard?
In this post, we’ll focus on the Traffic Overview tab in the dashboard. This tool has the potential to be extremely valuable for a several reasons:
- Current Client Competitive Analysis – See how competitors are performing and identify similar traffic gains/losses
- New Business Pitch Prep – Oftentimes an RFP comes across your desk but without analytics access, forcing you to make assumptions
- Industry Category Health Check – See how websites in a particular industry vertical are performing after a Google algorithm change
- Competitive Insights: Discover how users engage with top brands in a given vertical (i.e. are users highly social in the online college industry?).
Other features of the new beta dashboard include:
- Subdomains– Reveals subdomains of a given root domain with corresponding traffic share percentage for each subdomain.
- Traffic Sources– A more in-depth look at the traffic sources listed in the Overview tab, including referring sites, social networks, and search engines.
- Geo Distribution– Distribution of site traffic, segmented by country and by what % of traffic that each subdomain represents for the domain entered.
- Destination Sites– A list of sites visited after users leave a domain
Our Methodology
Our analysis is based on the core metrics listed in the Traffic Overview dashboard tab. Since we work in search, we decided to focus on some of the most standard SEO metrics:
- Visits
- Average Visit Duration
- Bounce Rate
- Traffic Sources
- Geo Distribution
Using first party data for two of our client’s sites, we compared these same metrics over the same period of time in both Google Analytics and SEMrush’s Traffic Analytics.
Some Issues That We Ran Into
Unlike Google Analytics, SEMrush lumps both paid and organic search together as one traffic source: “Search”. Additionally, SEMrush does not specify if the metrics we analyzed are only for search, or an aggregate of all traffic sources.
Parsing the Data
Despite inconsistencies in the two platforms’ labeling, we looked at the data side-by-side to compare our first party, Google Analytics (GA), data and SEMrush’s third party data.
To get the best results, we looked at both combined search (organic and paid) and overall site traffic to find out which, if either, lined up with SEMrush new tool.
The Results for Client #1
| Data Point | Search | Total |
| Visits | (209,824) | (1,366,224) |
| Avg. Visit Duration | -14% | 35% |
| Bounce Rate | -18% | -9% |
| % Traffic Source | -1.9% | NA |
| Geo Distribution | Way off | Way off |
- Visits – Search visits were closer to the SEMrush number, but were still off by over 200,000 visits. For total site visits, SEMrush was off by nearly 1.4M visits.
- Average Visit Duration – Looking at Search alone, SEMrush underreported by 14%, but looking at the aggregate Total number, SEMrush overestimated by 35%.
- Bounce Rate – the GA data for Total & Search visitor’s bounces were also inaccurate; 9% and 18% higher than the GA data, respectively.
- % Traffic Source – we looked at the Search numbers only in this analysis, because the other numbers were off by too much to make a comparison. The SEMrush % Traffic Source for Search was the most accurate SEMrush data aligned to the GA data with only a 1.9% difference. This was the smallest discrepancy in our entire analysis.
- Geo Distribution – The Geo distribution was way off for every account that we looked. This functionality appears to be the least dependable.
The Results for Client #2
| Data Point | Search | Total |
| Visits | 158% | -97% |
| Avg. Visit Duration | -96% | -97% |
| Bounce Rate | -31% | -87% |
| % Traffic Source | 100.0% | – |
| Geo Distribution | Way off | Way off |
SEMrush Traffic Analytics Review Results & Takeaways
- Visits – SEMrush was off by 97% for Total visits and is underreporting Search visits by 158%.
- Average Visit Duration – Search and Total were both underreported by 96-97%.
- Bounce Rate – Search and Total were both underreported by 31% & 87%, respectively.
- % Traffic Source – 100% of traffic was awarded to search, which is not accurate.
- Geo Distribution – also way off for this account, as well.
Our Conclusion
We love SEMRUSH and regularly use it for competitive analysis. That being said for this new tool we found the metrics were not consistent with GA, and in some cases were wildly inaccurate. It’s important to keep in mind that this dashboard is still in beta. We hope that this analysis can serve as useful feedback for the SEMrush team and aid them in the next version of this tool. Please let us know what you think in the comments.
Where is SEMrush getting its web analytics data?
SEMrush has about 100 different databases where they hold scraped SERP results, Adwords volume data and proprietary data parsing that ultimately produces, the, more-often-than-not, incorrect data. But as I mention in the article, flawed data is sometimes better than no data.